Meet the moderators

[Zoe7]

Ok - let's see how we go @Former-Member

[Former-Member]

Not complex sampling but:

As power decreses - what happens to type II errors?

1. Determining the sample interval (represented by k), randomly selecting a number between 1 and k, and including each kth element in your sample are the steps for which form of sampling?

		Systematic Sampling
		Simple Random Sampling
		Stratified Random Sampling
		Cluster sampling

1. Which statements are true about complex sampling involving clustering compared to simple random sampling?

		Complex sampling may increase logistic efficiency
		Complex sampling increases statistical efficiency
		Complex sampling decreases statistical efficiency
		To retain the same Confidence Interval Width as SRS you need to decrease the sample size

[Zoe7]

Off the top of my head @Former-Member - no chance of answering either question accurately but give me a little time to try and find some of my old books.

[Former-Member]

That's okay. 😊
What did you study again? Psych?

[Zoe7]

Arts first then Social Research then Teaching @Former-Member

[Former-Member]

That's a lot of study!
Yay for study! 🤓 📖 ☕️ 📝

[Zoe7]

It's been a long time since I have done any statistical work to this extent @Former-Member - from memory (which isn't that great atm ) I am pretty sure that Question 1 is Systemic Sampling as you are choosing the same random number in each interval. All the others (simple, stratified, cluster) use any random numbers generated/chosen across the whole sample being used without dividing the samples into equal/like intervals.

[Zoe7]

Yep @Former-Member it is alot - 7 years all up.. but I did go back 'later in life' to do teaching after working in a couple of different fields first.

[Zoe7]

@Former-Member Question 2: complex sampling involving clustering may 1. increase logistic efficiency and 2. decrease statistical efficiency

I am not 100% sure these are correct as it has been a long time since I have done statistical analysis to this depth but by making the sample groups 'chosen' smaller (into clusters) it makes sense.

[Former-Member]

Thanks @Zoe7.
I'm not sure I understand the answer..... more reading for me I guess!
Thank you.